The rules our reviews answer to
Our methodology explains how we test and score. This page is the layer above it: the principles that govern every review — who controls them, what standards they meet, how we use AI in our own work, and the firm line we keep between making money and reaching a verdict.
Last updated 26 June 2026One promise underpins everything here: the verdict belongs to the reader, not to an advertiser, a partner, or our own bottom line. We're a small independent publisher, and our only asset is whether you can trust the score. We protect that above traffic, above revenue, above any single tool's opinion of us.
The wall between money and verdicts
How we fund the site and how we judge a tool are two separate worlds. This is what lives on each side — and what is never allowed to cross.
How we keep the lights on
- Affiliate links that may earn a commission
- Display advertising, where present
- Partnerships and sponsorships, clearly labelled
How we reach a verdict
- Hands-on testing of every tool
- Six-dimension scores and the overall number
- Pros, cons, recommendations and rankings
- Payment for a score — no one can buy a higher number
- Payment for coverage — money doesn't decide what we review
- Vendor sign-off — no company approves a review before it runs
- Affiliate value — a bigger commission never lifts a verdict
Six principles every review follows
Evidence over assertion
Every claim is grounded in first-hand testing or aggregated real reviews — never in marketing copy or guesswork.
People first, not algorithms
We write for the person deciding whether to spend, not to please a search engine. Useful beats optimised.
Independence
Verdicts aren't for sale. No one outside our editorial team signs off, softens, or sets a score.
Fairness
Every tool is judged on the same six dimensions, and companies get a right of reply on matters of fact.
Transparency
We show the prompts and outputs, date every test, name the testers, and disclose how we make money.
Accountability
A named human owns every review and every correction. When we're wrong, we fix it openly .
How we use AI in our own reviews
We review AI tools, so it's reasonable to ask how much AI writes the reviews. Here's our line, plainly.
A human tests every tool, and the verdict is always a human's. AI can't sign up, pay, run a real task, or feel where a product is frustrating to use — so the experience and the judgement behind every score come from a person. AI may help draft, summarise or tidy language, but a human tests, verifies, edits, and is accountable for every published word.
AI may assist with drafting, organising notes, or tightening prose — under human direction.
A human verifies every fact — prices, limits and features are checked against the live tool.
We don't publish unverified AI output as fact, and AI never sets or changes a score.
We never fabricate quotes, user reviews, ratings or screenshots — with or without AI.
How money works here
The short version lives below; the full detail is in our advertising & affiliate disclosure .
What we do
- Use affiliate links that may earn a commission at no extra cost to you
- Disclose when a vendor provides free or trial access for testing
- Clearly label any sponsored or partner content and keep it apart from reviews
- Keep editorial judgement entirely separate from commercial relationships
What we never do
- Accept payment for a review, a higher score, or favourable coverage
- Let an affiliate commission influence what we recommend
- Give any company approval over a review before it's published
- Disguise an ad or a paid placement as an independent verdict
Sourcing, and fairness
Sourcing & attribution
Claims rest on our own hands-on testing and on user reviews aggregated from named platforms — set out in full in our methodology . We report patterns rather than outliers, attribute user sentiment to where it came from, and don't pass off anyone else's work as our own.
Fairness & right of reply
If a company believes a review contains a factual error, it can flag it through our corrections process , and we'll fix anything genuinely wrong. What we won't do is change an honest, tested verdict because a vendor disagrees with the outcome — a low score earned on evidence stays.